Water conservancy project data analysis method and system based on big data

Abstract

The embodiment of the invention provides a hydraulic engineering data analysis method and system based on big data, and belongs to the technical field of data analysis. The method comprises the following steps: obtaining corresponding initial hydrological data of a target object under target sensors at different positions at a target moment and obtaining an initial meteorological text related to the target object; performing state recognition on the initial hydrological data to obtain a target state corresponding to the target object; determining a first description keyword corresponding to the target object at the target moment according to a target state and initial hydrological data in combination with a word mapping table; performing semantic feature recognition on the initial meteorological text to obtain a second description keyword corresponding to the initial meteorological text; determining a deviation value between the initial hydrological data and the initial meteorological text of the target object at the target moment according to the first description keyword and the second description keyword; and determining a corresponding data analysis result of the target object at the target moment according to the deviation value.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to a water conservancy project data analysis method and system based on big data. Background Art

[0002] Due to the complexity of water conservancy projects and the dynamic changes in their environment, a single data source often cannot fully capture these changes, which may lead to limitations in data analysis results. In particular, if you rely solely on hydrological data, you may ignore other key factors, which in turn affects the reliability of data analysis. In the prior art, many studies and practices focus on analyzing water conservancy project data from the perspective of multiple hydrological data. Although this is an improvement over single hydrological data, it may still not be comprehensive enough. This approach may ignore the comprehensive analysis with other types of data, resulting in low reliability of data analysis results. Summary of the invention

[0003] The main purpose of the embodiments of the present invention is to provide a water conservancy project data analysis method and system based on big data, aiming to solve the problem in the related technology that data analysis in the field of water conservancy projects is not comprehensive enough, resulting in low reliability of data analysis results.

[0004] In a first aspect, an embodiment of the present invention provides a water conservancy project data analysis method based on big data, comprising:

[0005] Obtaining initial hydrological data corresponding to target sensors at different positions of a target object at a target time and obtaining initial meteorological text related to the target object;

[0006] Performing state recognition on the initial hydrological data to obtain a target state corresponding to the target object;

[0007] Determine a first description keyword corresponding to the target object at the target time according to the target state and the initial hydrological data combined with a word mapping table;

[0008] Performing semantic feature recognition on the initial meteorological text to obtain a second description keyword corresponding to the initial meteorological text;

[0009] Determine, according to the first description keyword and the second description keyword, a deviation value between the initial hydrological data and the initial meteorological text of the target object at the target time;

[0010] The data analysis result corresponding to the target object at the target time is determined according to the deviation value.

[0011] In a second aspect, an embodiment of the present invention provides a water conservancy project data analysis system based on big data, comprising:

[0012] A data acquisition module, used to obtain initial hydrological data corresponding to target sensors at different positions of a target object at a target time and to obtain initial meteorological text related to the target object;

[0013] A state recognition module, used to perform state recognition on the initial hydrological data to obtain a target state corresponding to the target object;

[0014] A word conversion module, used to determine a first description keyword corresponding to the target object at the target time according to the target state and the initial hydrological data in combination with a word mapping table;

[0015] A semantic recognition module, used for performing semantic feature recognition on the initial meteorological text to obtain a second description keyword corresponding to the initial meteorological text;

[0016] A deviation identification module, used for determining a deviation value between the initial hydrological data and the initial meteorological text of the target object at the target time according to the first description keyword and the second description keyword;

[0017] A result determination module is used to determine the data analysis result corresponding to the target object at the target time according to the deviation value.

[0018] The embodiment of the present invention provides a water conservancy project data analysis method and system based on big data, the method comprising: obtaining initial hydrological data corresponding to a target object under target sensors at different positions at a target time and obtaining initial meteorological text related to the target object; thereby, by comprehensively analyzing the initial hydrological data and the initial meteorological text, the limitation of a single data source is avoided, making the data analysis more comprehensive. The combination of hydrological data and meteorological text can provide richer information and help understand the complexity and variability of hydrological phenomena. Then, the initial hydrological data is state-identified to obtain the target state corresponding to the target object, and then the operation state of the water conservancy project can be monitored in real time, and dynamic changes can be captured, so that when the target state is a preset state, the first description keyword corresponding to the target object at the target time is determined according to the initial hydrological data combined with a word mapping table; then, the initial meteorological text is semantically identified to obtain a second description keyword corresponding to the initial meteorological text; thereby, the initial hydrological data and the initial meteorological text are converted into comparable keywords through the word mapping table and semantic feature recognition, reducing the subjectivity of human interpretation and improving the objectivity and accuracy of the analysis. Therefore, the deviation value between the initial hydrological data and the initial meteorological text of the target object at the target time can be determined according to the first description keyword and the second description keyword; the deviation between the initial hydrological data and the initial meteorological text can be quantified, so as to more accurately judge the difference and consistency between the data, avoiding the misleading of a single indicator, and finally determining the data analysis result corresponding to the target object at the target time according to the deviation value, thereby providing intelligent decision support for the operation and management of water conservancy projects. This method also solves the problem in related technologies that the data analysis in the field of water conservancy projects is not comprehensive enough, resulting in low reliability of data analysis results. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 A schematic diagram of a flow chart of a water conservancy project data analysis method based on big data provided by an embodiment of the present invention;

[0021] Figure 2 A schematic diagram of the module structure of a water conservancy project data analysis system based on big data provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0024] It should be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0025] The embodiment of the present invention provides a water conservancy project data analysis method and system based on big data. The water conservancy project data analysis method based on big data can be applied to a terminal device, which can be an electronic device such as a tablet computer, a laptop computer, a desktop computer, a personal digital assistant, and a wearable device. The terminal device can be a server or a server cluster.

[0026] Some embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0027] Please refer to Figure 1 , Figure 1 A schematic flow chart of a water conservancy project data analysis method based on big data provided in an embodiment of the present invention.

[0028] like Figure 1 As shown, the water conservancy project data analysis method based on big data includes steps S101 to S106.

[0029] Step S101: obtaining initial hydrological data corresponding to a target object at target sensors at different positions at a target time and obtaining initial meteorological text related to the target object.

[0030] Exemplarily, the target object may be a river, a reservoir or a canal, etc., and thus, according to the characteristics of the target object, sensors such as water level meters, flow meters, and rain gauges are arranged at different locations such as upstream, midstream, and downstream to monitor hydrological data in real time, so that the initial hydrological data corresponding to the target object at different locations can be obtained from the target sensor periodically or in real time. The initial hydrological data includes but is not limited to water level, flow, precipitation, etc.

[0031] Exemplarily, a meteorological data source related to the target object is selected, such as: meteorological reports issued by the National Meteorological Administration (such as weather forecasts, extreme weather warnings), real-time meteorological data provided by meteorological stations (such as temperature, humidity, wind speed, precipitation, etc.), satellite remote sensing data (such as cloud maps, precipitation distribution, etc.), and then the initial meteorological text corresponding to the target object is obtained from the meteorological data source through an API interface or crawler technology. The initial meteorological text is used to characterize the meteorological information that affects the water volume information corresponding to the target object. In other words, the initial meteorological text records various weather conditions and phenomena related to the water volume changes of the target object, such as precipitation, evaporation, runoff, etc., which directly affect the operating status of the target object.

[0032] Step S102: performing state recognition on the initial hydrological data to obtain a target state corresponding to the target object.

[0033] Exemplarily, machine learning or a traditional model (such as a classification model) is used to classify and identify the state of the initial hydrological data to obtain the target state corresponding to the target object. The target state is either a normal state of the target object or an abnormal state of the target object.

[0034] In some embodiments, the state identification of the initial hydrological data to obtain the target state corresponding to the target object includes: obtaining first hydrological data corresponding to a first position from the initial hydrological data, and performing abnormality identification on the first hydrological data to obtain an abnormal state corresponding to the first position; obtaining second hydrological data corresponding to a second position adjacent to the first position from the initial hydrological data based on the abnormal state; obtaining an abnormality confidence level corresponding to the first position under the abnormal state based on the first hydrological data and the second hydrological data; and determining the target state corresponding to the target object based on the abnormality confidence level and the first position.

[0035] Exemplarily, data related to the first position is screened out from the initial hydrological data. For example, if the first position is an upstream monitoring point of the target object, the hydrological data such as water level and flow of the upstream monitoring point are extracted to obtain the first hydrological data corresponding to the target object at the first position.

[0036] Exemplarily, the first hydrological data is judged to be abnormal by using statistical features such as the mean and standard deviation of historical data, or the abnormality is judged by analyzing the change trend of the first hydrological data to obtain the abnormal state corresponding to the first position. The abnormal state can be either the first position is normal or the first position is abnormal.

[0037] Exemplarily, when the abnormal state corresponding to the first position is that the first position is abnormal, the second position adjacent to the first position is obtained. For example, if the first position is an upstream monitoring point of a river, the second position may be a midstream monitoring point, and the second hydrological data of the second position at the target time is extracted from the initial hydrological data.

[0038] Exemplarily, a weight is allocated according to the importance of the first position and the second position, and a deviation between the first hydrological data and the second hydrological data is calculated, so as to determine the abnormal confidence corresponding to the first position in an abnormal state according to the weight information and the deviation information.

[0039] Exemplarily, according to the size of the abnormality confidence and the importance of the first position, a judgment rule for the target state is formulated. For example, if the abnormality confidence is high and the first position is a key monitoring point, the target state is "severe abnormality". If the abnormality confidence is medium, the target state is "potential abnormality". If the abnormality confidence is low, the target state is "normal".

[0040] In some embodiments, the abnormality identification of the first hydrological data to obtain the abnormal state corresponding to the first position includes: determining a target window, and segmenting the first hydrological data according to the target window to obtain a plurality of window hydrological data; performing difference calculation on the continuous hydrological data in each window hydrological data to obtain a difference hydrological sequence corresponding to the window hydrological data; performing mean and variance calculation on the difference hydrological sequence to obtain a window mean and a window variance; determining a confidence interval corresponding to the window hydrological data according to the window mean and the window variance, and determining a mean estimate corresponding to the window hydrological data according to the confidence interval. The method comprises the steps of: performing extreme value processing on all the mean estimation intervals to determine the upper and lower mean limits corresponding to the first hydrological data; performing standard deviation calculation on each of the window hydrological data to obtain the standard deviation information corresponding to the window hydrological data and obtaining the mathematical expectation corresponding to each of the window hydrological data; determining the expected data range corresponding to the window hydrological data according to the standard deviation information, the mathematical expectation, the upper and lower mean limits; performing abnormality identification on the window hydrological data according to the expected data range to obtain the degree of abnormality corresponding to the window hydrological data; and determining the abnormal state corresponding to the first position according to the degree of abnormality.

[0041] Exemplarily, the target window is determined based on expert experience or historical experience, and then the first hydrological data is segmented according to the target window to obtain multiple window hydrological data. For example, the first hydrological data is segmented by sliding using the target window according to the target step length to obtain multiple window hydrological data.

[0042] Exemplarily, the continuous hydrological data in each window of hydrological data are subjected to difference calculation according to adjacent data points, and then the difference results are combined into a difference hydrological sequence according to the difference calculation order. The mean and variance of the difference hydrological sequence are then calculated to obtain the window mean and window variance.

[0043] Exemplarily, assuming that the difference hydrological sequence obeys a normal distribution, since the population variance corresponding to the window hydrological data is unknown, the target variance and target mean corresponding to the difference hydrological sequence can be used to obtain the confidence interval of the population data mean corresponding to the window hydrological data that obeys a normal distribution, and then the confidence level can be determined. The mean of the window hydrological data can be estimated based on the confidence level and confidence interval to obtain the mean estimation interval.

[0044] For example, assuming that the difference hydrological sequence follows a normal distribution and the population variance corresponding to the window hydrological data is unknown, the t distribution can be used to calculate the confidence interval of the population mean. Specifically, assume that the difference hydrological sequence follows a normal distribution. This is usually based on the central limit theorem, that is, the sum of a large number of independent random variables approximately follows a normal distribution. For the difference hydrological sequence of each target window, the window mean and window variance are calculated. Since the population variance is unknown, the t distribution is used to calculate the confidence interval. Thus, using the above confidence interval, the mean estimation interval of each target window is obtained, which represents the possible range of the population mean under a certain confidence level.

[0045] Exemplarily, extreme value processing is performed on the mean estimation interval of the window hydrological data corresponding to all target windows to obtain the maximum value of the upper limit of the mean estimation interval of all window hydrological data and the minimum value of the lower limit of the mean estimation interval of all window hydrological data, thereby obtaining the upper limit and lower limit of the mean of the first hydrological data.

[0046] Exemplarily, the standard deviation of each window hydrological data is calculated using a statistical formula to obtain the standard deviation information corresponding to the window hydrological data, and the mathematical expectation corresponding to each window hydrological data is obtained according to a statistical algorithm, thereby determining the expected data range corresponding to the window hydrological data based on the standard deviation information, the mathematical expectation, the upper limit of the mean, and the lower limit of the mean.

[0047] For example, the data volume corresponding to the first hydrological data is obtained, and the data volume is squared to obtain the square root value, and then the standard deviation information is divided by the square root value to obtain the target value, and then the target value is multiplied by the lower limit of the mean to obtain the first value, and the target value is multiplied by the upper limit of the mean to obtain the second value, and then the mathematical expectation is subtracted from the first value to obtain the minimum value corresponding to the expected data range, and the mathematical expectation is added to the second value to obtain the maximum value corresponding to the expected data range, and then the minimum value and maximum value corresponding to the expected data range are obtained.

[0048] Exemplarily, the window hydrological data is compared with the expected data range to obtain the abnormal hydrological quantity corresponding to the window hydrological data not being within the expected data range, and then the degree of abnormality corresponding to the window hydrological data is determined based on the ratio between the abnormal hydrological quantity and the data volume corresponding to the window hydrological data.

[0049] For example, based on historical data or industry standards, multiple preset value ranges are set, and each range corresponds to a different degree of abnormality. For example, slight abnormality, medium abnormality, and severe abnormality. For the window hydrological data of each target window, the ratio between the number of abnormal hydrological data and the amount of data corresponding to the window hydrological data is calculated, and then the calculated ratio is compared with the preset value range to determine in which preset value range the ratio falls. Thus, according to the preset value range in which the ratio is located, the degree of abnormality of the window hydrological data is determined accordingly. For example, the preset value range ratio is defined as a slight abnormality within the range of [a, b); the abnormality within the range of [b, c) is a medium abnormality; the abnormality within the range of [c, ∞) is a severe abnormality, and so on.

[0050] Exemplarily, the abnormality levels of all target windows are comprehensively analyzed. For example, if most target windows show severe abnormality, the abnormal state of the first position is severe abnormality. If a few target windows show slight abnormality, the abnormal state of the first position is normal.

[0051] In some embodiments, the obtaining of the abnormality confidence corresponding to the first position in the abnormal state according to the first hydrological data and the second hydrological data includes: calculating the distance information between the first position and the second position, and determining the weight information of the second position to the first position according to the distance information; performing data distribution analysis on the first hydrological data to obtain first distribution information corresponding to the first hydrological data and performing data distribution analysis on the second hydrological data to obtain second distribution information corresponding to the second hydrological data; determining the information similarity value between the first position and the second position according to the first distribution information and the second distribution information; obtaining the number of positions with the same information distribution as the first position from the second position according to the information similarity value; determining the abnormality confidence corresponding to the first position in the abnormal state according to the number of positions, the information similarity value and the weight information; wherein the abnormality confidence is obtained according to the following formula:

[0052]

[0053] in, represents the abnormality confidence corresponding to the i-th first position; represents the number of positions corresponding to the i-th first position, represents the weight information of the j-th second position having the same information distribution as the first position to the i-th first position, represents the number of positions corresponding to the jth second position having the same information distribution as the first position; Represents the information similarity value between the i-th first position and the j-th second position having the same information distribution as the first position.

[0054] Exemplarily, geographic coordinates (such as longitude and latitude) are used to calculate the distance between the first position and the second position using the Haversine formula to obtain distance information between the first position and the second position, and then obtain distance information between the first position and all second positions, and then the reciprocal of all distance information is taken separately to obtain reciprocal information, and then all reciprocal information is summed to obtain a target sum value, and then the reciprocal information between the first position and the second position is divided by the target sum value to obtain weight information of the second position to the first position.

[0055] Exemplarily, a statistical method is used to perform data distribution analysis on first hydrological data at a first location to obtain first distribution information corresponding to the first hydrological data, and a statistical method is used to perform data distribution analysis on second hydrological data at a second location to obtain second distribution information corresponding to the second hydrological data.

[0056] Exemplarily, the information entropy between the first distribution information and the second distribution information is calculated using Kullback-Leibler divergence, so as to determine the information similarity value between the first position and the second position according to the information entropy.

[0057] Exemplarily, a preset similarity value is determined, and then the preset similarity value and the information similarity value are compared. Therefore, when the information similarity value is greater than or equal to the preset similarity value, it is determined that the information distribution between the second distribution information corresponding to the second position corresponding to the information similarity value and the first distribution information corresponding to the first position is the same, thereby obtaining the number of positions corresponding to the second position when the information similarity value is greater than or equal to the preset similarity value.

[0058] Exemplarily, the abnormality confidence of the first position in the abnormal state is calculated according to the following formula in combination with the number of positions, the information similarity value and the weight information:

[0059]

[0060] in, represents the abnormal confidence corresponding to the i-th first position; represents the number of positions corresponding to the i-th first position, represents the weight information of the j-th second position having the same information distribution as the first position to the i-th first position, represents the number of positions corresponding to the jth second position having the same information distribution as the first position; Represents the information similarity value between the i-th first position and the j-th second position having the same information distribution as the first position.

[0061] Exemplarily, by considering the number of second locations that have the same information distribution as the first location, it is possible to evaluate whether the anomaly occurs in a larger area, thereby increasing the confidence, and then assigning weights based on the distance between the locations. The closer the distance, the greater the impact of the second location on the first location, and therefore the higher the weight, thereby making the evaluation of the anomaly confidence more comprehensive and accurate, and then according to the above formula, the anomaly confidence of the first location in an abnormal state can be effectively determined.

[0062] Step S103: determining a first description keyword corresponding to the target object at the target time according to the target state and the initial hydrological data in combination with a word mapping table.

[0063] Exemplarily, when the target state of the monitored target object is in the target abnormality, the identified abnormal trend corresponding to the initial hydrological data is obtained. The abnormal trend refers to the dynamic evolution process of the data deviating from the normal mode, which can reflect the direction and speed of the abnormal phenomenon over time. At the same time, it is also necessary to obtain the abnormal distribution related to the target abnormality. The abnormal distribution describes the distribution characteristics of the abnormal phenomenon in space, such as the frequency and intensity of the abnormal occurrence and its performance in different locations. After obtaining the information of the abnormal trend and abnormal distribution, the word mapping table is used to convert these complex technical data into easy-to-understand descriptive language, that is, the first description keyword. The word mapping table defines the correspondence between different types of abnormal trends and abnormal distributions and descriptive keywords. For example, if the abnormal trend shows that the data is rising rapidly, and the abnormal distribution reveals that the abnormal phenomenon is concentrated in a specific area, then the corresponding description keywords may be "sharp rise" and "local concentration".

[0064] Step S104: performing semantic feature recognition on the initial meteorological text to obtain a second description keyword corresponding to the initial meteorological text.

[0065] Exemplarily, a graph-based keyword extraction algorithm, similar to the PageRank algorithm, extracts keywords from the initial meteorological text to obtain the second description keywords corresponding to the initial meteorological text. Alternatively, a topic model is used to extract keywords from the initial meteorological text to obtain the second description keywords corresponding to the initial meteorological text. Alternatively, a combination of methods can achieve better results. For example, TF-IDF can be used to initially extract candidate keywords, and then the quality of keywords can be further optimized through TextRank or BERT. By comprehensively considering word frequency, semantics, and contextual information, the second description keywords corresponding to the initial meteorological text can be more accurately extracted.

[0066] In some embodiments, the performing semantic feature recognition on the initial meteorological text to obtain the second description keyword corresponding to the initial meteorological text includes: performing a preprocessing operation on the initial meteorological text to obtain an initial word segmentation result corresponding to the initial meteorological text; obtaining an association relationship between any two word segmentation data in the initial word segmentation result based on a positional relationship of the initial word segmentation result in the initial meteorological text; obtaining a directly associated participle and an indirectly associated participle corresponding to any one of the word segmentation data in the initial word segmentation result based on the association relationship; determining an importance score corresponding to the word segmentation data based on the directly associated participle and the indirectly associated participle; and obtaining the second description keyword corresponding to the initial meteorological text from the initial word segmentation result based on the importance score.

[0067] Exemplarily, preprocessing operations include but are not limited to text cleaning such as removing irrelevant characters, punctuation marks, special symbols, etc., and operations including using appropriate word segmentation tools (such as Jieba, NLTK, etc.) to segment the initial meteorological text into words or vocabulary units, thereby obtaining an initial word segmentation result, that is, a list of all word segments in the initial meteorological text.

[0068] Exemplarily, it is determined whether any two segmentation data in the initial segmentation result are in the same sentence in the initial meteorological text, so that when any two segmentation data are in the same sentence in the initial meteorological text, the association relationship between the two segmentation data is determined to be a direct connection.

[0069] For example, in order to determine whether the word segmentation data are in the same sentence, the initial meteorological text needs to be segmented into multiple sentences first. The basis for sentence segmentation is usually punctuation marks (such as periods, semicolons, exclamation marks, etc.). Next, each word segmentation data is assigned the sentence number in which it is located, so that through this mapping, it is clear which sentence each word segmentation data is in. If two word segmentation data are in the same sentence, it is considered that there is a direct connection between them, that is, they are semantically related. Then for any two word segmentation data, check whether their sentence numbers are the same. If any two word segmentation data are checked and their sentence numbers are the same, it is determined that the association relationship between the two word segmentation data is that there is a direct connection.

[0070] For example, in order to more intuitively represent the association relationship between word segmentation data, an association relationship graph can be constructed. Each node in the graph represents a word segmentation data. If there is a direct connection between two word segmentation data, an edge is used to connect the two nodes to obtain an association relationship graph.

[0071] Exemplarily, the directly associated participles and indirectly associated participles corresponding to any participle data in the initial participle results are obtained according to the association relationship graph, wherein the directly associated participles represent data that are directly related to the participle data; and the indirectly associated participles represent data that are further connected to the participle data through the directly associated participles.

[0072] Exemplarily, PageRank, TextRank and other algorithms are used to calculate the importance of indirect correlation segmentation and direct correlation analysis in the initial meteorological text, and then the importance score corresponding to the segmentation data is obtained. For example, the PageRank value that has comprehensively considered the direct correlation segmentation and the indirect correlation segmentation is directly used as the importance score.

[0073] Exemplarily, a threshold is set, and segmented data with importance scores exceeding the threshold are selected or the first N segmented data with the highest importance scores are selected to be determined as the second description keywords corresponding to the initial meteorological text.

[0074] In some embodiments, determining the importance score corresponding to the segmentation data according to the directly associated segmentation and the indirectly associated segmentation includes: calculating a first similarity between the directly associated segmentation and the segmentation data and calculating a second similarity between the indirectly associated segmentation and the segmentation data; determining an association gap value between the indirectly associated segmentation and the segmentation data according to the second similarity; determining the importance score corresponding to the segmentation data according to the first similarity and the association gap value; wherein the importance score is obtained according to the following formula:

[0075]

[0076] in, represents the importance score corresponding to the t-th word segmentation data, represents the number of segmentations corresponding to the directly associated segmentation corresponding to the t-th segmentation data, represents the first similarity between the t-th word segmentation data and the s-th directly related word segmentation data; Indicates the number of participles corresponding to the indirect related participle corresponding to the t-th participle data; represents the second similarity between the t-th word segmentation data and the h-th indirectly related word segmentation data; Represents the association gap value between the tth word segmentation data and the hth indirectly related word segmentation data.

[0077] Exemplarily, a similarity measurement method is used, such as cosine similarity, Jaccard similarity, word vector distance, etc., to calculate the first similarity between each directly associated segmentation and the segmentation data. A similarity measurement method similar to the measurement method of direct similarity is used to calculate the second similarity between each indirectly associated segmentation and the segmentation data.

[0078] Exemplarily, the reciprocal of the second similarity is determined as the association gap value between the indirectly associated segmented word and the segmented word data. The association gap value reflects the degree of difference in describing the topic between the indirectly associated segmented word and the segmented word data.

[0079] Exemplarily, the importance score of the word segmentation data is determined using the first similarity and the associated gap value according to the following formula:

[0080]

[0081] in, Indicates the importance score corresponding to the t-th word segmentation data, Indicates the number of segmentations corresponding to the directly associated segmentation corresponding to the t-th segmentation data, Represents the first similarity between the tth word segmentation data and the sth directly related word segmentation; Indicates the number of participles corresponding to the indirect related participles corresponding to the t-th participle data; represents the second similarity between the tth word segmentation data and the hth indirectly related word segmentation; Represents the correlation gap value between the tth word segmentation data and the hth indirectly related word segmentation data.

[0082] For example, the similarity of the segmentation data directly related to the segmentation data is considered through the first similarity, reflecting the importance of the segmentation data in the local context. The difference of the segmentation data indirectly related to the segmentation data is considered through the association gap value, reflecting the degree of difference in the description of the topic between the indirectly related segmentation and the segmentation data. Therefore, by combining the similarity information of the directly related segmentation and the indirectly related segmentation, the semantic importance of the segmentation data can be captured more accurately, thereby improving the accuracy of keyword extraction.

[0083] Step S105: determining a deviation value between the initial hydrological data and the initial meteorological text of the target object at the target time according to the first description keyword and the second description keyword.

[0084] Exemplarily, similarity calculation is performed on the first description keyword and the second description keyword to obtain information similarity of the target object between the initial hydrological data and the initial meteorological text, thereby determining the deviation value between the initial hydrological data and the initial meteorological text according to the information similarity.

[0085] In some embodiments, determining the deviation value between the initial hydrological data and the initial meteorological text of the target object at the target time based on the first description keyword and the second description keyword includes: calculating the third similarity between each second keyword in the second description keyword and each first keyword in the first description keyword; obtaining the target weight corresponding to the second keyword from the initial meteorological text; and determining the deviation value between the initial hydrological data and the initial meteorological text based on the target weight and the third similarity.

[0086] Exemplarily, a semantic similarity measurement method, such as cosine similarity, Jaccard similarity, word vector distance, etc., is used to calculate the third similarity between each second keyword in the second description keywords and each first keyword in the first description keywords.

[0087] Exemplarily, TF-IDF is used to determine the target weight corresponding to the second keyword obtained from the initial meteorological text, and then the weighted similarity between each second keyword and all the first keywords is calculated based on the target weight and the third similarity, so that the weighted similarities of all the second keywords are combined to determine the target similarity between the initial hydrological data and the initial meteorological text, and then the deviation value between the initial hydrological data and the initial meteorological text is determined based on the target similarity.

[0088] In some embodiments, determining the deviation value between the initial hydrological data and the initial meteorological text based on the target weight and the third similarity includes: determining the associated keyword corresponding to the second keyword in the first description keyword based on the third similarity; obtaining the fourth similarity between the second keyword and the associated keyword from the third similarity; obtaining the maximum similarity between the second keyword and the associated keyword from the fourth similarity; and determining the deviation value between the initial hydrological data and the initial meteorological text based on the target weight and the maximum similarity.

[0089] Exemplarily, for each second keyword, the third similarity between the second keyword and all the first keywords is calculated using cosine similarity, and the first keyword with the highest third similarity is selected as the associated keyword corresponding to the second keyword.

[0090] Exemplarily, a fourth similarity between each second keyword and its corresponding associated keyword is obtained from the third similarity, and then the maximum similarity between the second keyword and the associated keyword is obtained from the fourth similarity. Thus, the maximum similarities corresponding to all second keywords are summed to obtain a similarity sum value, and then the target weight corresponding to each second keyword is divided by the similarity sum value to obtain a target ratio, and then all target ratios are summed to obtain a deviation value between the initial hydrological data and the initial meteorological text.

[0091] Exemplarily, the maximum similarity is extracted through the fourth similarity to reflect the semantic closeness between the second keyword and the associated keyword. This helps to more finely measure the relevance between keywords, and then through the combination of target weight and maximum similarity, comprehensively reflect the impact of the second description keyword on the deviation value. The introduction of target weight makes the calculation of the deviation value more consider the importance of keywords in the text.

[0092] Step S106: determining the data analysis result corresponding to the target object at the target time according to the deviation value.

[0093] Exemplarily, the deviation value is mapped to the state interval of the target object (e.g., abnormal, normal, slight deviation, etc.). For example: a positive deviation value may indicate that the initial hydrological data is higher than expected, corresponding to an "abnormally high value". A negative deviation value may indicate that the initial hydrological data is lower than expected, corresponding to an "abnormally low value". A deviation value close to zero may indicate that the hydrological data is basically consistent with expectations, corresponding to "normal". Thus, after obtaining the state interval corresponding to the deviation value, the data analysis result corresponding to the target object at the target time is determined according to the state information corresponding to the deviation value.

[0094] In some embodiments, determining the data analysis result corresponding to the target object at the target moment according to the deviation value includes: determining a preset value, and comparing the preset value and the deviation value to obtain a target comparison result; when the target comparison result is that the deviation value is greater than or equal to the preset value, then determining that the target judgment result of the target object at the target moment is target abnormality, and then determining the data analysis result corresponding to the target object at the target moment according to the target abnormality; when the target comparison result is that the deviation value is less than the preset value, then determining that the target judgment result of the target object at the target moment is target normal, and then determining the data analysis result corresponding to the target object at the target moment according to the target normality.

[0095] Exemplarily, a preset value is set based on historical data or expert experience to determine whether the state of the target object at the target time is abnormal, and then the calculated deviation value is compared with the preset value. If the deviation value is greater than or equal to the preset value, the target comparison result is "the deviation value is greater than or equal to the preset value". If the deviation value is less than the preset value, the target comparison result is "the deviation value is less than the preset value".

[0096] Exemplarily, when the deviation value is greater than or equal to the preset value, the target judgment result of the target object at the target time is determined to be target abnormal, which means that the data of the target object at the target time deviates from the normal range, and there may be risks or further analysis is required. When the deviation value is less than the preset value, the target judgment result of the target object at the target time is determined to be target normal. Target normal means that the data of the target object at the target time is within the normal range and no special attention is required.

[0097] Exemplarily, when the target judgment result is that the target is abnormal, it is necessary to analyze the possible causes of the target abnormality based on the initial hydrological data and the initial meteorological text, such as drastic changes in meteorological conditions, abnormal fluctuations in initial hydrological data, etc., and then obtain the data analysis result: "The target object is abnormal at the target time, the deviation value is X, and the possible causes include Y and Z. Further verification is recommended." When the target judgment result is that the target is normal, the data analysis result is: "The target object is in a normal state at the target time, the deviation value is X, which is in line with expectations." That is, a risk warning or abnormality report is generated when the target is abnormal, and a normal state report is generated when the target is normal.

[0098] See also Figure 2 , Figure 2 A water conservancy project data analysis system 200 based on big data is provided in an embodiment of the present application. The water conservancy project data analysis system 200 based on big data includes a data acquisition module 201, a state recognition module 202, a word conversion module 203, a semantic recognition module 204, a deviation recognition module 205, and a result determination module 206. The data acquisition module 201 is used to obtain the initial hydrological data corresponding to the target object at the target time under the target sensor at different positions and to obtain the initial meteorological text related to the target object; the state recognition module 202 is used to perform state recognition on the initial hydrological data to obtain the target state corresponding to the target object. state; a word conversion module 203, used to determine the first description keyword corresponding to the target object at the target time according to the target state in combination with the word mapping table; a semantic recognition module 204, used to perform semantic feature recognition on the initial meteorological text to obtain the second description keyword corresponding to the initial meteorological text; a deviation recognition module 205, used to determine the deviation value between the initial hydrological data and the initial meteorological text of the target object at the target time according to the first description keyword and the second description keyword; a result determination module 206, used to determine the data analysis result corresponding to the target object at the target time according to the deviation value.

[0099] In some implementations, the state identification module 202 performs, during the process of performing state identification on the initial hydrological data to obtain the target state corresponding to the target object:

[0100] Obtaining first hydrological data corresponding to a first position from the initial hydrological data, and performing abnormality identification on the first hydrological data to obtain an abnormal state corresponding to the first position;

[0101] Obtaining second hydrological data corresponding to a second position adjacent to the first position from the initial hydrological data according to the abnormal state;

[0102] Obtaining an abnormality confidence corresponding to the first position under the abnormal state according to the first hydrological data and the second hydrological data;

[0103] The target state corresponding to the target object is determined according to the abnormality confidence and the first position.

[0104] In some implementations, the state identification module 202, in the process of identifying anomalies in the first hydrological data to obtain an abnormal state corresponding to the first location, executes:

[0105] Determine a target window, and perform data segmentation on the first hydrological data according to the target window to obtain hydrological data of multiple windows;

[0106] Performing difference calculation on the continuous hydrological data in each window of hydrological data to obtain a difference hydrological sequence corresponding to the window of hydrological data;

[0107] Calculating the mean and variance of the difference hydrological series to obtain the window mean and window variance;

[0108] Determine a confidence interval corresponding to the window hydrological data according to the window mean and the window variance, and determine a mean estimation interval corresponding to the window hydrological data according to the confidence interval;

[0109] Performing extreme value processing on all the mean estimation intervals to determine the mean upper limit and mean lower limit corresponding to the first hydrological data;

[0110] Performing standard deviation calculation on each of the window hydrological data to obtain standard deviation information corresponding to the window hydrological data and obtaining mathematical expectations corresponding to each of the window hydrological data;

[0111] Determine the expected data range corresponding to the window hydrological data according to the standard deviation information, the mathematical expectation, the upper limit of the mean and the lower limit of the mean;

[0112] Performing abnormal identification on the window hydrological data according to the expected data range to obtain the abnormal degree corresponding to the window hydrological data;

[0113] The abnormal state corresponding to the first position is determined according to the abnormality degree.

[0114] In some implementations, the state identification module 202, in the process of obtaining the abnormality confidence corresponding to the first position in the abnormal state according to the first hydrological data and the second hydrological data, performs:

[0115] Calculating distance information between the first position and the second position, and determining weight information of the second position to the first position according to the distance information;

[0116] Performing data distribution analysis on the first hydrological data to obtain first distribution information corresponding to the first hydrological data and performing data distribution analysis on the second hydrological data to obtain second distribution information corresponding to the second hydrological data;

[0117] Determine an information similarity value between the first position and the second position according to the first distribution information and the second distribution information;

[0118] According to the information similarity value, the number of positions with the same information distribution as the first position is obtained from the second position;

[0119] Determine the abnormality confidence corresponding to the first position in the abnormal state according to the number of positions, the information similarity value and the weight information;

[0120] The abnormality confidence is obtained according to the following formula:

[0121]

[0122] in, represents the abnormality confidence corresponding to the i-th first position; represents the number of positions corresponding to the i-th first position, represents the weight information of the j-th second position having the same information distribution as the first position to the i-th first position, represents the number of positions corresponding to the jth second position having the same information distribution as the first position; Represents the information similarity value between the i-th first position and the j-th second position having the same information distribution as the first position.

[0123] In some implementations, the semantic recognition module 204 performs, during the process of performing semantic feature recognition on the initial meteorological text to obtain the second description keyword corresponding to the initial meteorological text:

[0124] Performing a preprocessing operation on the initial meteorological text to obtain an initial word segmentation result corresponding to the initial meteorological text;

[0125] Obtaining the association relationship between any two word segmentation data in the initial word segmentation result according to the position relationship of the initial word segmentation result in the initial meteorological text;

[0126] Obtaining, according to the association relationship, a directly associated participle and an indirectly associated participle corresponding to any one of the participle data in the initial participle result;

[0127] Determine the importance score corresponding to the segmentation data according to the directly associated segmentation and the indirectly associated segmentation;

[0128] The second description keyword corresponding to the initial meteorological text is obtained from the initial word segmentation result according to the importance score.

[0129] In some implementations, the semantic recognition module 204, in the process of determining the importance score corresponding to the segmentation data according to the directly associated segmentation and the indirectly associated segmentation, performs:

[0130] Calculating a first similarity between the directly associated segmented word and the segmented word data and calculating a second similarity between the indirectly associated segmented word and the segmented word data;

[0131] Determine the association gap value between the indirectly associated word segmentation and the word segmentation data according to the second similarity;

[0132] Determining the importance score corresponding to the word segmentation data according to the first similarity and the associated gap value;

[0133] The importance score is obtained according to the following formula:

[0134]

[0135] in, represents the importance score corresponding to the t-th word segmentation data, represents the number of segmentations corresponding to the directly associated segmentation corresponding to the t-th segmentation data, represents the first similarity between the t-th word segmentation data and the s-th directly related word segmentation data; Indicates the number of participles corresponding to the indirect related participle corresponding to the t-th participle data; represents the second similarity between the t-th word segmentation data and the h-th indirectly related word segmentation data; Represents the association gap value between the tth word segmentation data and the hth indirectly related word segmentation data.

[0136] In some implementations, the deviation identification module 205, in the process of determining the deviation value between the initial hydrological data and the initial meteorological text of the target object at the target time according to the first description keyword and the second description keyword, performs:

[0137] calculating a third similarity between each second keyword in the second description keywords and each first keyword in the first description keywords;

[0138] Obtaining a target weight corresponding to the second keyword from the initial meteorological text;

[0139] The deviation value between the initial hydrological data and the initial meteorological text is determined according to the target weight and the third similarity.

[0140] In some implementations, the deviation identification module 205, during the process of determining the deviation value between the initial hydrological data and the initial meteorological text according to the target weight and the third similarity, performs:

[0141] Determine, according to the third similarity, an associated keyword corresponding to the second keyword in the first description keyword;

[0142] Obtaining a fourth similarity between the second keyword and the associated keyword from the third similarity;

[0143] Obtaining the maximum similarity between the second keyword and the associated keyword from the fourth similarity;

[0144] The deviation value between the initial hydrological data and the initial meteorological text is determined according to the target weight and the maximum similarity.

[0145] In some implementations, during the process of determining the data analysis result corresponding to the target object at the target time according to the deviation value, the result determination module 206 executes:

[0146] Determine a preset value, and compare the preset value with the deviation value to obtain a target comparison result;

[0147] When the target comparison result is that the deviation value is greater than or equal to the preset value, it is determined that the target judgment result of the target object at the target moment is a target abnormality, and then the data analysis result corresponding to the target object at the target moment is determined according to the target abnormality;

[0148] When the target comparison result is that the deviation value is less than the preset value, the target judgment result of the target object at the target moment is determined to be target normal, and then the data analysis result corresponding to the target object at the target moment is determined based on the target normal.

[0149] In some embodiments, the water conservancy project data analysis system 200 based on big data can be applied to terminal devices.

[0150] It should be noted that technical personnel in the relevant field can clearly understand that, for the convenience and conciseness of description, the specific working process of the water conservancy project data analysis system 200 based on big data described above can refer to the corresponding process in the aforementioned water conservancy project data analysis method based on big data embodiment, and will not be repeated here.

[0151] An embodiment of the present invention also provides a storage medium for computer-readable storage, wherein the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of any one of the water conservancy project data analysis methods based on big data provided in the description of the embodiment of the present invention.

[0152] The storage medium may be an internal storage unit of the terminal device described in the foregoing embodiment, such as a hard disk or memory of the terminal device. The storage medium may also be an external storage device of the terminal device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc., equipped on the terminal device.

[0153] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In a hardware embodiment, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or transient medium). As is known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0154] It should be understood that the term "and / or" used in the present specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, including these combinations. It should be noted that, in this article, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "including a..." does not exclude the presence of other identical elements in the process, method, article or system including the element.

[0155] The serial numbers of the embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments. The above description is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A water conservancy project data analysis method based on big data, characterized in that: The method comprises: Obtaining initial hydrological data corresponding to target sensors at different positions of a target object at a target time and obtaining initial meteorological text related to the target object; Performing state recognition on the initial hydrological data to obtain a target state corresponding to the target object; Determine a first description keyword corresponding to the target object at the target time according to the target state and the initial hydrological data combined with a word mapping table; Performing semantic feature recognition on the initial meteorological text to obtain a second description keyword corresponding to the initial meteorological text; Determine, according to the first description keyword and the second description keyword, a deviation value between the initial hydrological data and the initial meteorological text of the target object at the target time; The data analysis result corresponding to the target object at the target time is determined according to the deviation value.

2. The method according to claim 1, characterized in that The performing state identification on the initial hydrological data to obtain the target state corresponding to the target object includes: Obtaining first hydrological data corresponding to a first position from the initial hydrological data, and performing abnormality identification on the first hydrological data to obtain an abnormal state corresponding to the first position; Obtaining second hydrological data corresponding to a second position adjacent to the first position from the initial hydrological data according to the abnormal state; Obtaining an abnormality confidence corresponding to the first position under the abnormal state according to the first hydrological data and the second hydrological data; The target state corresponding to the target object is determined according to the abnormality confidence and the first position.

3. The method according to claim 2, characterized in that The performing abnormality identification on the first hydrological data to obtain an abnormal state corresponding to the first position includes: Determine a target window, and perform data segmentation on the first hydrological data according to the target window to obtain hydrological data of multiple windows; Performing difference calculation on the continuous hydrological data in each window of hydrological data to obtain a difference hydrological sequence corresponding to the window of hydrological data; Calculating the mean and variance of the difference hydrological series to obtain the window mean and window variance; Determine a confidence interval corresponding to the window hydrological data according to the window mean and the window variance, and determine a mean estimation interval corresponding to the window hydrological data according to the confidence interval; Performing extreme value processing on all the mean estimation intervals to determine the mean upper limit and mean lower limit corresponding to the first hydrological data; Performing standard deviation calculation on each of the window hydrological data to obtain standard deviation information corresponding to the window hydrological data and obtaining mathematical expectations corresponding to each of the window hydrological data; Determine the expected data range corresponding to the window hydrological data according to the standard deviation information, the mathematical expectation, the upper limit of the mean and the lower limit of the mean; Performing abnormal identification on the window hydrological data according to the expected data range to obtain the abnormal degree corresponding to the window hydrological data; The abnormal state corresponding to the first position is determined according to the abnormality degree.

4. The method according to claim 2, characterized in that: The obtaining, according to the first hydrological data and the second hydrological data, the abnormality confidence corresponding to the first position in the abnormal state includes: Calculating distance information between the first position and the second position, and determining weight information of the second position to the first position according to the distance information; Performing data distribution analysis on the first hydrological data to obtain first distribution information corresponding to the first hydrological data and performing data distribution analysis on the second hydrological data to obtain second distribution information corresponding to the second hydrological data; Determine an information similarity value between the first position and the second position according to the first distribution information and the second distribution information; According to the information similarity value, the number of positions with the same information distribution as the first position is obtained from the second position; Determine the abnormality confidence corresponding to the first position in the abnormal state according to the number of positions, the information similarity value and the weight information; The abnormality confidence is obtained according to the following formula: in, represents the abnormality confidence corresponding to the i-th first position; represents the number of positions corresponding to the i-th first position, represents the weight information of the j-th second position having the same information distribution as the first position to the i-th first position, represents the number of positions corresponding to the jth second position having the same information distribution as the first position; Represents the information similarity value between the i-th first position and the j-th second position having the same information distribution as the first position.

5. The method according to claim 1, characterized in that The performing semantic feature recognition on the initial meteorological text to obtain a second description keyword corresponding to the initial meteorological text includes: Performing a preprocessing operation on the initial meteorological text to obtain an initial word segmentation result corresponding to the initial meteorological text; Obtaining the association relationship between any two word segmentation data in the initial word segmentation result according to the position relationship of the initial word segmentation result in the initial meteorological text; Obtaining, according to the association relationship, a directly associated participle and an indirectly associated participle corresponding to any one of the participle data in the initial participle result; Determine the importance score corresponding to the segmentation data according to the directly associated segmentation and the indirectly associated segmentation; The second description keyword corresponding to the initial meteorological text is obtained from the initial word segmentation result according to the importance score.

6. The method according to claim 5, characterized in that The determining the importance score corresponding to the segmentation data according to the directly associated segmentation and the indirectly associated segmentation includes: Calculating a first similarity between the directly associated segmented word and the segmented word data and calculating a second similarity between the indirectly associated segmented word and the segmented word data; Determine the association gap value between the indirectly associated word segmentation and the word segmentation data according to the second similarity; Determining the importance score corresponding to the word segmentation data according to the first similarity and the associated gap value; The importance score is obtained according to the following formula: in, represents the importance score corresponding to the t-th word segmentation data, represents the number of segmentations corresponding to the directly associated segmentation corresponding to the t-th segmentation data, represents the first similarity between the t-th word segmentation data and the s-th directly related word segmentation data; Indicates the number of participles corresponding to the indirect related participle corresponding to the t-th participle data; represents the second similarity between the t-th word segmentation data and the h-th indirectly related word segmentation data; Represents the association gap value between the tth word segmentation data and the hth indirectly related word segmentation data.

7. The method according to claim 1, characterized in that The step of determining the deviation value between the initial hydrological data and the initial meteorological text of the target object at the target time according to the first description keyword and the second description keyword includes: calculating a third similarity between each second keyword in the second description keywords and each first keyword in the first description keywords; Obtaining a target weight corresponding to the second keyword from the initial meteorological text; The deviation value between the initial hydrological data and the initial meteorological text is determined according to the target weight and the third similarity.

8. The method according to claim 7, characterized in that The step of determining the deviation value between the initial hydrological data and the initial meteorological text according to the target weight and the third similarity includes: Determine, according to the third similarity, an associated keyword corresponding to the second keyword in the first description keyword; Obtaining a fourth similarity between the second keyword and the associated keyword from the third similarity; Obtaining the maximum similarity between the second keyword and the associated keyword from the fourth similarity; The deviation value between the initial hydrological data and the initial meteorological text is determined according to the target weight and the maximum similarity.

9. The method according to claim 1, characterized in that: The determining, according to the deviation value, a data analysis result corresponding to the target object at the target time includes: Determine a preset value, and compare the preset value with the deviation value to obtain a target comparison result; When the target comparison result is that the deviation value is greater than or equal to the preset value, it is determined that the target judgment result of the target object at the target moment is a target abnormality, and then the data analysis result corresponding to the target object at the target moment is determined according to the target abnormality; When the target comparison result is that the deviation value is less than the preset value, the target judgment result of the target object at the target moment is determined to be target normal, and then the data analysis result corresponding to the target object at the target moment is determined based on the target normal.

10. A water conservancy project data analysis system based on big data, characterized in that: include: A data acquisition module, used to obtain initial hydrological data corresponding to target sensors at different positions of a target object at a target time and to obtain initial meteorological text related to the target object; A state recognition module, used to perform state recognition on the initial hydrological data to obtain a target state corresponding to the target object; A word conversion module, used to determine a first description keyword corresponding to the target object at the target time according to the target state and the initial hydrological data in combination with a word mapping table; A semantic recognition module, used for performing semantic feature recognition on the initial meteorological text to obtain a second description keyword corresponding to the initial meteorological text; A deviation identification module, used for determining a deviation value between the initial hydrological data and the initial meteorological text of the target object at the target time according to the first description keyword and the second description keyword; A result determination module is used to determine the data analysis result corresponding to the target object at the target time according to the deviation value.

Images